Update app.py

app.py

@@ -14,11 +14,11 @@ st.set_page_config(page_title="Chatbot Test", page_icon="🤖", layout="centered
 if "messages" not in st.session_state:
     st.session_state["messages"] = []
 
-# Function to Stream Response
-def stream_response(prompt_text, api_key):
+# Function to Fetch Response (Non-Streaming)
+def fetch_response(prompt_text, api_key):
     """
-    Stream text from the HF Inference Endpoint using the InferenceClient.
-    Yields each partial chunk of text as it arrives.
+    Fetch full text response from the HF Inference Endpoint using the InferenceClient.
+    Returns the generated text as a whole.
     """
     client = InferenceClient(SPACE_URL, token=api_key)
 
@@ -31,15 +31,11 @@ def stream_response(prompt_text, api_key):
         "stop_sequences": ["<|endoftext|>"]
     }
 
-    stream = client.text_generation(prompt_text, stream=True, details=True, **gen_kwargs)
-
     try:
-        for response in stream:
-            if response.token.special:
-                continue
-            yield response.token.text
+        response = client.text_generation(prompt_text, stream=False, details=True, **gen_kwargs)
+        return response["generated_text"]
     except Exception as e:
-        yield f"Error: {e}"
+        return f"Error: {e}"
 
 # Streamlit Chat Interface
 st.title("Chatbot Testing Interface")
@@ -60,13 +56,8 @@ if prompt:
     # 3) Generate the response
     with st.chat_message("assistant", avatar=DUBS_PATH):
         with st.spinner("Dubs is thinking... Woof Woof! 🐾"):
-            full_response = ""
-            placeholder = st.empty()  # Placeholder for streaming response
-            response = stream_response(chat_history, HF_API_KEY)
-            for item in response:
-                full_response += item
-                placeholder.markdown(full_response)
-            placeholder.markdown(full_response)
+            full_response = fetch_response(chat_history, HF_API_KEY)
+            st.write(full_response)
 
     # 4) Add assistant response to the session state
     st.session_state["messages"].append({"role": "assistant", "content": full_response})